Embodiments of the present invention relate generally to apparatus and methods for differentiating constituent types of formed bodies (e.g., cells or other small particles) in liquid samples (e.g., whole blood or other particle-suspending liquids). More particularly, embodiments relate to improvements in transducer assemblies for use in hematology analyzers and other flow cytometers that function to sense, characterize, and differentiate formed bodies in such samples by various optical parameters, often in combination with non-optical parameters acquired via the Coulter Principle. Embodiments of the invention further relate to improvements in optical flow cells used in such transducer assemblies and to improvements in methods for sensing characterizing parameters from formed bodies transiting the parameter-acquisition portion of such flow cells.
The analyses of patient body fluids can be automated as an aid in diagnosing a patient's state of health. Such analyses can include flowing a prepared portion of such body fluids through a transducer assembly to derive certain parameters characteristic of the several different types or subpopulations of constituent formed bodies therein, differentiating and enumerating the several types or subpopulations of the formed bodies on the basis of the derived parameters, and processing or correlating the resultant information to provide desired diagnostics. For example, these tasks can be accomplished for whole blood via characterization of blood cells therein by automated hematology analyzers and flow cytometers.
A fundamental performance limit of some current instruments of the type noted above originates in the transduction of formed-body properties into the characterizing parameters used to assign individual formed bodies to specific subpopulations. For many applications, optical transductive methods alone provide effective means for characterizing formed bodies. In these applications, a portion of a prepared sample is interrogated with optical radiation as it flows through a passageway formed in an optically transparent element, or flow cell, forming a portion of a transducer module. Suitable photo-detectors, also forming part of the module, are positioned to detect various optical parameters from an irradiated formed body including, for example, its optical absorbance of the interrogating beam, its fluorescence at different wavelengths, and its light-scattering effect within one or more angular ranges. In these optical-only applications, it will be appreciated that the physical extent of the flow-cell passageway can be relatively large without adversely affecting the determination of these optical parameters. However, in other cytometric applications where such optical parameters are combined with simultaneously determined non-optical parameters, in particular those based on the Coulter Principle (discussed below), both the cross-section and at least a length of a portion of the flow-cell passageway may need to be dramatically restricted to achieve suitable signal strength.
Due to the limitations of then-existing optical sensing methods, W. H. Coulter devised an electronic method for characterizing minute formed bodies suspended in a liquid. The now-familiar Coulter Principle enables determination of the volume of formed bodies, by flowing a sample portion prepared in an electrically conductive liquid through a particle-sensing zone (or volumeter conduit) simultaneously with an electric current. The electrical resistivity of the particle-suspending liquid differs from that of the particles, the electrical contrast permitting counting and sizing of particles transiting the volumeter conduit. Although other geometries were discussed in Coulter's U.S. Pat. No. 2,656,508, volumeter conduits are usually cylindrical bores in a thin insulative wafer, the conduit's cross-sectional area and length determining volumetric sensitivity, coincidence volume, and maximum passable formed-body dimension; thus, conduit diameters are typically at most an order of magnitude greater than the diameter of the typical formed bodies to be analyzed. The volumeter conduit forms the only fluidic communication between two insulative chambers of the transducer assembly, with no requirement on the optical characteristics of the wafer material surrounding the conduit. Initially, a direct current (DC) was provided through the conduit, and resistive Coulter volume (V) signals proportional to the volume of transiting formed bodies were acquired via electrodes positioned outside the opposite ends of the volumeter conduit. In U.S. Pat. No. 3,380,584 differences in the Coulter V signals were adapted to sort subpopulations of such formed bodies to separate collection sites. In U.S. Pat. No. 3,502,974 to W. H. Coulter and W. R. Hogg, the excitation current through the conduit was made to include at least one alternating current (AC), thereby permitting determination of not only the resistive but also reactive components of the conduit current resulting from its modulation by passage of a formed body. When such currents include one having a frequency in the radio-frequency (RF) range (e.g., 22.5 MHz), the respective components permit estimation of the volume (V) and electrical conductivity (C) of a formed body, and the ratio of the reactive to resistive components is said to be the “opacity” of the formed body.
In commonly assigned U.S. Pat. No. 5,125,737 to C. M. Rodriguez and W. H. Coulter (hereinafter, the '737 patent) the Coulter V and C parameters are combined with optical scatter (S) parameters. In certain hematology analyzers incorporating this VCS technology, a diluted blood sample is passed through a cytometric transducer assembly that includes an optical flow cell made from a short segment of optically transparent cylindrical tubing. It is known in the glass-working art to draw tubing from a larger preform of various inner and outer diameters and having a cylindrical wall up to several centimeters thick. During the drawing operation, the preform is heated to a predetermined temperature at which its viscosity permits deformation, whereupon it is drawn axially, usually in a vertically downward direction, at a constant and predetermined rate. During this process the diameters of the inner and outer cross-sections of the preform are substantially reduced with the original circular shapes being substantially retained, such shapes being the minimum energy shape, and the wall of the preform is significantly reduced in thickness. This drawing process has been adapted to form thick-wall transparent tubing for use in producing seamless single-piece (monolithic) flow cells such as described in the '737 patent. After a preform is drawn to a preferred inner diameter (e.g., 50 micra), the tubing is cut into segments of a preferred length (e.g., about 6 mm), and a suitable flat (e.g., 1 mm wide) is lapped and polished on the exterior cylindrical surface of the segments which, as a result of the drawing, has been reduced to a diameter of about 3.5 mm. The flat provides an optical port through which a radiation beam (e.g., from a HeNe laser) can be coupled, substantially perpendicular thereto and along a diameter, to the central cylindrical channel in the drawn tube. The flat is made substantially parallel to the channel axis at an arbitrary location on the outer cylindrical surface, e.g., to avoid or eliminate an objectionable optical defect in the segment wall. The cylindrical channel is partially enlarged by boring a segment from both ends to a suitable diameter (e.g., about 1.2 mm), leaving in situ a short length (e.g., about 65 micra) of the original channel in the middle of the flow cell that opens at each end into a cup-shaped recess (e.g., of radius about 600 micra) substantially coaxial with the original channel and continuous with the end bores. In use, the length of the original channel in the longitudinal section thus formed functions as a Coulter volumeter conduit, and the cylindrical bores of the resulting passageway communicate with external electrode chambers in said transducer assembly, whereby the sample liquid in a hydrodynamically-focusing sheath liquid can be made to pass centrally through the volumeter conduit. Coulter V and C parameters are acquired from electrodes as described in aforesaid U.S. Pat. Nos. 2,656,508 and 3,502,974 as formed bodies pass through the volumeter conduit; meanwhile, optical scatter (S) parameters from the radiation beam are simultaneously acquired from the individual formed bodies as they pass seriatim through said beam traversing the parameter sensing zone formed by the cylindrical volumeter conduit.
In certain hematology analyzers, the repeatable circular cross-section of the flow-cell volumeter conduit is of designed dimension, and the flatted cylindrical surface forming the envelope of the '737 flow cell is substantially parallel to the axis of the conduit. The incoming radiation beam is brought through the flat to a minimum cross-sectional area within the circular volumeter conduit. Light from the beam is scattered by passing nucleated cells in the sample, and photo-detectors suitably positioned near the optical axis detect the forward-scattered radiation within specific angular bands, thereby allowing the aforesaid S parameters to be developed. Such forward-scatter (FS) signals acquired through the wall of such flow cells permit reliable enumeration, differentiation, and classification of normal individual leukocytes into monocytes, lymphocytes, neutrophils, eosinophils, and basophils when appropriately correlated with Coulter V and C parameters. These diagnostic data have a proven history of clinical usefulness.
At interfaces between dissimilar transparent materials an incident radiation ray of wavelength λ, the normal to the interfacial tangent at the point of the ray's incidence, and the ray emerging from the interface are all co-planar. The ray paths follow Snell's Law, n1(λ) sin θ1=n2(λ) sin θ2, where θ1 and θ2 are the respective angles at which a ray is incident on the interface and emerges from it, both with respect to said normal, and n1(λ) and n2(λ) are the refractive indices on the incident and exit sides of the interface. Its low refractive index [n(λ)≈1.457] led to fused silica (SiO2) being preferred for use in aforesaid '737 flow cells. For typical suspending liquids the refractive index is about 1.333, while that of air is 1.000. For any wall geometry in fused silica and sin θ2 less than unity, rays originating near the flow-cell axis will be refracted toward the surface normal at their incidence on the conduit surface (sin θ2≈0.915 sin θ1) but away from it at their incidence on the envelope surface (sin θ2≈1.457 sin θ1). Scattered radiation propagating from formed bodies within the '737 volumeter conduit to the relevant photo-detectors passes through suspending liquid in contact with the cylindrical conduit surface and the cylindrical envelope surface in contact with ambient air, the cylindrical wall between said surfaces thus being a non-axisymmetric refractive element. For rays originating on the axis of '737 flow cells and exiting the volumeter conduit in the plane including the optical axis of the incoming radiation beam, θ1=θ2=0, and such rays pass through both surfaces of the flow-cell wall without significant refraction. However, if n1(λ)≠n2(λ) such rays propagating at any angle θ to said plane will be refracted at both wall surfaces, the refractive deviation from the path of incidence increasing with the combined effects of θ and the mismatch between n1(λ) and n2(λ) with the refraction being symmetric about said plane. Thus, a circular cone of rays exiting through the flow-cell wall from an origin at the intersection of the optical and conduit axes will be asymmetrically refracted, with refractive deviation ranging from zero where said plane intersects either the conduit or envelope surface to total internal reflection at the envelope surface if θ1 makes 1.457 sin θ1≥1. Due to these non-axisymmetric refractive effects, scattered radiation passing through the flow-cell wall acquires substantial astigmatism that, affecting scatter from small objects more than that from larger objects, affects the ability to differentiate formed bodies characterized by granular structure. Parallel planar wall surfaces minimize such asymmetric refractive effects, and such a cone of rays experiences uniform refraction about the optical axis of the radiation beam as determined by the angle of incidence and the difference between n1(λ) and n2(λ) across both wall surfaces. Such walls improve acquisition of not only S but other optical signals, and extensive effort has been directed toward flow cells having a prismatic channel surrounded by planar walls, e.g., a modified embodiment in the '737 patent incorporates a flow cell that includes a square volumeter conduit within a similar envelope.
In commonly assigned U.S. Pat. No. 6,228,652 to C. M. Rodriguez et al. (hereinafter, the '652 patent), experimental apparatus is disclosed that can provide simultaneous acquisition of various optical, Coulter V, and Coulter C signals from an individual formed body, with subsequent differentiation of formed-body subpopulations in whole blood based thereon. One of the square flow-cell structures illustrated in the '737 patent is the preferred flow cell in the '652 patent and is discussed regarding FIG. 3 therein. This flow cell comprises an optically transparent element having a prismatic exterior envelope of square cross-section, measuring about 4.2 mm on each side, and having a length of about 6.3 mm. (As used hereinafter, the word “prismatic” refers to any three-dimensional figure composed of three or more intersecting sides that are planar, and a pair of opposing ends that are polygonal in shape. Hereinafter, “polygonal” is used to refer to any closed plane figure having at least three substantially straight sides, and “planar” as used herein refers to a surface having an area that is predominantly flat.) Centrally located within said prismatic element is a prismatic volumeter conduit having a square cross-section about 50 micra on each side and a length of about 65 micra; the relatively small cross-section and length of the conduit are necessary to attain a reasonable volumetric sensitivity and coincidence volume for acquiring said V and C signals. Thus, the ratio of the respective cross-sectional areas of said conduit and envelope is approximately 0.00014, and the wall thickness is about 2.075 mm. To acceptably limit aberrational content of optical signals, surfaces of the prismatic envelope and conduit should be substantially parallel, with optical planarity. This combination of square/square cross-sectional geometries, wall thickness, wall surface parallelism, and wall flatness is difficult to achieve, a difficulty compounded by the small dimensions required for volumeter conduits.
To manufacture flow cells of the type preferred in the '652 patent, a relatively complex planarization process has been used wherein four transparent plates, e.g., preferably made of a form of silica, are polished to predetermined thickness and finish and assembled to form the composite structure of FIG. 1. During assembly, a pair of said plates CC1 and CC3 is spaced apart by the other pair CC2 and CC4 to form two walls of flow cell 20, with the pair of equal-thickness spacer plates appropriately spaced apart a predetermined distance so that their opposing edges complete an internal channel 22. Preferably, complementary elements CC1-CC4 are of appropriate dimensions and joined at their interfaces by fusion to form a prismatic rod having an internal, longitudinally-extending, straight channel 22 of a desired uniform square cross-section therein. This rod is then cut to a desired length, e.g., 6.3 mm, and the segments polished to the desired external geometry and dimensions to form a flow cell having a prismatic envelope, e.g., opposing planar sides 50 in FIG. 1 having a flat-to-flat separation of 4.2 mm. Such composite optical flow cells having prismatic flow channels of constant longitudinal section but various square cross-sections have been made by varying the thickness and separation of the two spacer plates CC2 and CC4. For acquisition of Coulter parameters, a passageway including a prismatic volumeter conduit, e.g., the 50 micra by 50 micra square conduit described above, is formed in such flow cells by boring square channel 22 from both ends as described above for the '737 flow cell to form a parameter-acquisition zone. The longitudinal section of said passageway is such that the sample liquid in a hydrodynamically-focusing sheath liquid passes centrally through the square volumeter conduit thus formed within the flow cell. Spacer plates may also be separated by a spacing differing from their thickness, to form flow channels having rectangular cross-sections as in U.S. Pat. No. 4,786,165. Various embodiments of flow cells made by the planarization process have been adapted to function in certain flow cytometers. Flow cells suited to application in such instruments and made by the method of the present invention are illustrated in respective FIGS. 7A, 7B, 12, and 14 and will be discussed as embodiments of the present invention.
Although useful flow cells of the type preferred in the '652 patent have been produced by the above-described planarization process, the yield of such flow cells processed to include a volumeter conduit is very low, typically less than 1 in 3, due to weakness in the fused joins. For formed bodies transiting the internal volumeter conduit of useful flow cells (e.g., BC2 in FIG. 1), optical signals acquired through the two windows may be substantially repeatable (e.g., the forward-scatter signals FS resulting from a sensor placed outside window CC1 and on the optical axis OA opposite the entry window CC3 for radiation beam B in FIG. 1). However, those acquired through the two walls of flow cell 20 containing spacer plates (e.g., side-scatter signal SS acquired through CC2 and fluorescence signal F acquired through CC4 in FIG. 1) demonstrate both sensitivity to excitation beam position in individual cases and unit-to-unit variability in the resulting optical signals, even though joins are systematically positioned relative to the axis of optical excitation. And because liquid flows and Coulter excitation currents through the flow-cell passageway contact the exposed assembly joins, flow cells made by the planarization process (e.g., flow cell 20 in FIG. 1) are also prone to subsequent failure modes: Firstly, irregularities and air pockets in the fused joins characteristic of the planarization method result in localized heating due to the RF component of the conduit excitation current, with consequent failure of the join; secondly, join imperfections tend to enlarge when sample flows are accompanied by significant cyclic pressure; and finally, flow cells left in stored apparatus tend to separate along the joins due to crystallization of salts if residual reagents are allowed to evaporate.
More-complex production processes, wherein various transparent solids of predetermined geometry are appropriately assembled, have also been used to make composite optical flow cells having prismatic volumeter conduits of a desired geometry and dimensions. If the flow-cell wall were to consist of a spherical envelope containing a concentric spherical cavity, a cone of rays originating at the intersection of the optical and conduit axes would be propagated with no refraction; a flow-cell wall including a spherical envelope centered on the axis of the volumeter conduit enables propagation of scattered radiation with minimal wall-induced refractive aberrations to the extent that n1(λ) approximates n2(λ) at the surface of the conduit. Thus, commonly assigned U.S. Pat. No. 4,348,107 (hereinafter, the '107 patent) discloses optical flow cells in which a volumeter conduit having a preferably square cross-section is contained within an envelope having an exterior spherical surface or other surface of revolution. (Flow cells having such envelope, but made by method embodiments of the present invention, are illustrated in FIG. 13A-13C and will be discussed as embodiments of the present invention.) As illustrated in the '107 patent, such flow cells are made by joining together four complementary, truncated, square-based pyramids formed of a transparent material. The apex is polished from each pyramid parallel to its base and to a depth calculated to yield one side of the desired volumeter conduit, and the pyramids then appropriately assembled and adhesively joined together so that the truncated apexes form an unobstructed square prismatic conduit, the adjacent faces of the joined pyramids forming a tapering longitudinal section at one or both ends of the conduit. Although optical signals may be acquired through the planar surfaces of the resulting prismatic envelope, it is preferred that means unspecified in the '107 patent then provide the flow cell an envelope formed as a surface of revolution. An extension allowing coupling of the sample liquid and a hydrodynamically-focusing sheath liquid through the conduit is sealed to the resulting flow cell where its surface is intersected by one or both of the approaches formed by the exposed sides of the four pyramids. The '107 patent notes that optical and mechanical characteristics of said structure proved suboptimal, the adhesive joins potentially fluorescing or separating, but provides no alternative joining method. A theoretical comparison, of the optical properties of the idealized '107 flow-cell structure with those of a flow cell having a square cross-section in a square prismatic envelope, was published in Applied Optics (26:3244-3248, 1987) by the inventor and one of the present co-inventors; no method for production of either flow-cell structure was described. The inventor and other co-workers later verified some of those theoretical predictions in a comparison (Cytometry 20:185-190, 1995) of an embodiment of the '107 flow cell, one having the four pyramids fused together and thus avoiding said problems of adhesive joins, to a monolithic cylindrical flow cell as discussed above regarding certain hematology analyzers incorporating VCS technology; FIG. 2 in said publication shows the '107 flow cell sealed, after production of a polished spherical exterior envelope, between extensions of the plastic chambers housing electrodes enabling acquisition of Coulter V and C signals. As disclosed in U.S. Pat. Nos. 4,673,288 and 4,818,103, variations of the approach disclosed in the '107 patent have been used to provide prismatic volumeter conduits having a triangular cross-section in a similarly shaped envelope, with square, five-sided, etc., structures said to be within the scope of the invention. To allow efficient collection by a microscope objective of optical signals from such triangular volumeter conduits, in U.S. Patent Application 2007/0085997 a thin transparent plate (window) is substituted for one of the truncated pyramids, with the envelope completed by the remaining two complementary components modified to facilitate interrogation of formed bodies by optical radiation through their walls. As will be appreciated, multiple joins, exposed to conduit contents and subject to the disadvantages described above for the planarization process, are required in such composite flow cells comprising a plurality of such fused elements. Further, tolerances in machining apexes from multiple elements, in assembly of the elements, and in joining them to form a volumeter conduit combine to produce variable conduit geometry and dimensions, making this approach both costly and unattractive as a production process and one yielding unit-to-unit variability in the result.
Other disadvantages originate in imperfections in the exposed joins used to assemble prior-art composite flow cells such as described in the '652 patent or in U.S. Patent Application 2007/0085997 and its precursors. Such joins have been made via adhesive processes, low-temperature glass-bonding processes using chemical agents or solder glasses, or high-temperature fusion processes in which surfaces of the complementary components to be joined are placed in close proximity and heated sufficiently to cause those surfaces to soften and bond to each other. Joins formed by the first two methods are significantly less durable than those formed by fusion of the complementary components and may result in background fluorescence that interferes with the weak fluorescent radiation emitted by formed bodies transiting the resulting parameter-acquisition zone; in addition, bonding agents may extend or leave a residue beyond the machined surfaces intended to define corner geometry of the sensing zone and thus cause unpredictable unit-to-unit variability in liquid flow through the zone. Alternatively, insufficient bonding agent may inadequately fill the gap between adjacent components of the composite flow cell, leaving a void extending between such components along the length of the corner these components were intended to form. Viscous forces acting on adjacent surfaces of non-cylindrical flow channels combine such that fluid flows near the corners experience additional resistance, and so slower flow velocities, than near the mid-portions of the surfaces. Consequently, formed bodies outside near-axial flow in non-cylindrical channels experience lower flow rates and may migrate into the corners of such flow channels (e.g., during flow transitions required for flushing of one sample from, and introduction of a different sample into, the transducer assembly). Due to the small dimensions of volumeter conduits and potential issues related to thoroughly flushing them, certain challenges may arise when Coulter V and/or C parameters are acquired and involve additional complexity in practical instrumentation. Typical formed bodies are at most several micra in dimension and thus can be sequestered in such interstices in imperfect joins during such transitions. On resumption of continuous flow the viscosity-induced low flow rates near channel corners may be insufficient to sweep all such sequestered cells out, allowing the potential carry-over of formed bodies from one sample into a subsequent sample. If such a sequestered formed body were of the rare cell types critical to diagnosis, it would not only be absent from the first sample, but could occur in a following normal sample. Misleading diagnostic information could result from the subsequent processing of parameters acquired from both samples as a result of carryover of formed bodies from one patient sample into another patient sample. Because of the difficulty in flushing the small volumeter conduits required for sensing Coulter V and/or C parameters, instrumentation using prior-art composite flow cells including fusion joins is subject to the latter fault and its implications as well. It was noted above regarding the monolithic cylindrical flow cells that these were readily formed because the channel shape was the minimum-energy shape for the glass when softened for drawing. Minimum-energy considerations also apply during fusion-joining of complementary components and result in rounding of the intersection of the surfaces to be joined and the surfaces intended to form the non-cylindrical flow channel. Thus, for example, in flow cells such as used in the experimental instrumentation described in the '652 patent (i.e., flow cell 20 in FIG. 1), the edges of the two spacer plates soften and round over about a center within the spacer plates before the bulk glass softens sufficiently to bond the surfaces to be joined; such rounding is indicated for one such edge of CC4 by R′ in FIG. 1, but applies to both such edges of spacer plates CC2 and CC4. As indicated in FIG. 1, the cross-section of resulting channel 22 is not truly rectilinear, but rather has imposed at said four corners interstices adjacent to the two window plates CC1 and CC3 and extending back from the intended corner for several micra. Interstices extending more than 15 micra away from the flow channel and along much of the channel length have been observed in commercial planarized flow cells used during development of said '652 instrumentation; these have a perceptible radius of several micra at both corners of both spacer plates CC2 and CC4. Such join interstices are subject to the aforesaid risk for carryover of formed bodies, with attendant regulatory and liability concerns.
Requiring no joins, truly monolithic flow cells made from a single piece of transparent material, e.g., flow cells of the aforesaid '737 construction, surround the liquid flows and any Coulter excitation currents with a joinless homogeneous wall and so can avoid the limitations and disadvantages described above for certain flow cells comprising joined components of complementary geometries. The single-piece design of such flow cells yields a robust sensing element providing both controlled geometry and dimensions in the flow channel and reliable function in service. When used in data acquisition as described in the '737 patent, forward-scatter (FS) signals acquired from such flow cells permit reliable differentiation and enumeration of individual formed bodies when suitably correlated with Coulter V and C signals. But as discussed above, due to the necessarily small diameters of their circular volumeter conduits, the wall of the '737 flow cells acts as a non-axisymmetric refractive element, with greater refraction for scatter from small objects than for larger objects, and scatter (S) signals acquired through it incur substantial astigmatism that limit the ability to differentiate between certain types of formed bodies. While such differentiation can be improved by adding fluorescence (F) signals at different wavelengths as in the '652 patent, e.g., by selectively tagging the formed bodies with fluorescent dyes or dye-bead conjugates, dispersion due to n(λ) causes F signals to be even more disadvantageously affected by wall-induced refractive artifacts than are scatter signals. Composite flow cells having a prismatic parameter-acquisition zone enclosed by planar wall surfaces (e.g., the square volumeter conduit disclosed in the '652 patent or the triangular volumeter conduit disclosed in U.S. Patent Application 2007/0085997 and its precursors) can also minimize dispersion effects. As noted, however, such flow cells are difficult to make by conventional production methods; exposed joins required for assembly not only introduce optical inhomogeneities, but have the potential in service for both carry-over of formed bodies between samples and failure over time.
A recurring cytometric need is to simultaneously acquire several different types of optical signals resulting from interaction of the formed bodies with one or more radiation sources, i.e., some combination of axial light-loss (ALL) signals; scatter (S) signals such as forward-scatter (FS), side-scatter (SS), or back-scatter (BS) signals; and multiple-wavelength fluorescence (F) signals. In such applications the three or four envelope surfaces on certain flow cells (i.e., those described in respective U.S. Patent Application 2007/0085997 and its precursors or the '652 patent) require that a plurality of sensors view the parameter-acquisition zone through complex beam-splitting and/or wavelength differentiating optics which, in addition to adding cost, introduce alignment and other optical difficulties. The need for more optical sensing paths might be addressed by adding planar surfaces to the flow-cell envelope (i.e., as illustrated in Japanese Unexamined Patent Application No. 62-168033) whereby the envelope could have a pentagonal, hexagonal, heptagonal, etc., cross-section, so that each optical measurement of interest could be made through a separate surface of the envelope. As implicit in discussion above of Snell's Law, such flow cells having a polygonal envelope and a cylindrical parameter-acquisition zone can reduce asymmetric refractive effects below those experienced with flow cells in which both envelope and sensing zone have cylindrical surfaces. Experimental flow cells, made by forming additional planar surfaces in appropriate spatial relation to the optical port lapped onto the drawn '737 flow cell, were found to reduce asymmetric refractive effects for some optical characterizing parameters acquired through said surfaces but, due to the cylindrical surface of the small-diameter sensing zone, not sufficiently for desired acquisition of other such parameters. In addition, further unit-to-unit variability in optical performance resulted from difficulty in aligning such planar surfaces with the drawn flow channel. Conversely, if the flow channel were made to comprise planar surfaces by assembly of complementary components, the manufacture of such composite flow cells would be impractically complex, time-consuming, and costly for commercial incorporation into cytometric instruments, and optically inhomogeneous joins may constrain such designs by incursion on a desired pattern of light collection. Further, reliability and liability implications would result during service, due to the aforesaid contact of the joins by the operational contents of the flow channel.
U.S. Pat. No. 8,189,187 to Graham et. al., (hereinafter, the '187 patent) discloses various embodiments and applications of monolithic optical flow cells formed from a prismatic flow cell, i.e., a monolithic structure made of silicon dioxide by glass-forming methods and having a through channel formed during that process suitable for containing cells in a fluid stream, said channel being defined by at least three substantially planar surfaces and of sufficient length as to permit measurement of cell characteristics by cytometric methods. Such prismatic flow cells are purchased having an as-drawn substantially cylindrical envelope coaxial with the prismatic through-channel, only said through-channel and a thicker wall distinguishing them geometrically from the aforesaid thick-wall transparent tubing used in producing monolithic flow cells as described regarding the '737 patent. To avoid the non-axisymmetric refraction inherent to such cylindrical envelopes, after receipt such prismatic flow cells are improved by providing them, via secondary machining processes, with an integral non-cylindrical envelope, thereby producing the monolithic optical flow cells of the '187 patent. Such monolithic flow cells comprising an envelope of square cross-section coaxial and parallel with a portion of prismatic interior channel of similar cross-section (i.e., a version of flow cell 20′ in FIG. 2) are used, according to other teachings of the '187 patent. Within the Coulter volumeter conduit Z so formed, Coulter V and C parameters as well as optical forward-scatter (FS) signals at multiple angles and axial light-loss (ALL) signals are acquired without the aforesaid functional limitations inherent in prior-art composite flow cells such as illustrated in the '737 patent and described in the '652 patent (e.g., flow cell 20 in FIG. 1) or disclosed in U.S. Patent Application 2007/0085997 and its several precursors. These cytometric characterization parameters enable improved discrimination of formed bodies, and the clinical value of diagnostics provided by these analyzers is now recognized. Further, the optical multi-port capability of Japanese Unexamined Patent Application No. 62-168033 is provided without the non-axisymmetric refraction inherent to its cylindrical flow channels. It has been found, however, that potential optical variability arising in the fabrication methods for monolithic flow cells may require selection of the resulting product at an advanced processing stage and so may limit the functional yield of flow cells that provide acceptable optical characterization parameters.
Prismatic flow cells are made via glass-forming methods, more fully described in the '187 patent, in which a relatively large cylindrical glass preform having an oversize internal prismatic channel of a desired polygonal cross-section is heated to a predetermined temperature at which its viscosity permits deformation and drawn axially on a conventional drawing tower to reduce the channel to a desired cross-sectional area. The necessary preform wall thickness is attained by sliding over, heating to cause a viscosity permitting deformation, collapsing onto, and fusing to a first silica tube, caused to have the desired channel cross-section by heating to cause a viscosity permitting deformation and collapsing said tube onto a mandrel having the desired channel geometry, a second larger cylindrical tube of appropriate inner and outer diameters (a sleeve tube) so as to seamlessly increase the wall thickness of the preform. Such oversleeving step is repeated with additional sleeve tubes of appropriate increasing inner and outer diameters until the preform wall thickness will provide a desired flow-cell wall thickness after the preform is drawn to yield the desired channel cross-sectional area and the drawn preform is machined to yield the desired envelope of a monolithic flow cell. Each of the aforesaid several tubes is preferably a form of silica (SiO2), most preferably synthetic amorphous silica, and each of the several heat cycles produces a viscosity in the range between 60×106 and 1×106 poise, more preferably between 28×106 and 3×106 poise. Such oversleeving process may result in a reduced yield of functional flow cells via four potential artifacts that may arise in one or more of the oversleeving cycles: a) air bubbles may be entrapped between the growing preform and the next sleeve tube, subsequently being drawn into air lines in a wall between the flow channel and an envelope surface (e.g., respectively Z and 50 in FIG. 2) of a finished flow cell that may interfere with acquisition of optical parameters; b) concentricity of the successive outer surfaces of the growing preform with the channel axis may be lost, whereby optical paths originating at the intersection of the channel and optical axes encounter refractive profiles through the wall of a finished flow cell that depend upon their angle with respect to the optical axis; c) the growing preform may soften sufficiently that wall surfaces of the internal flow channel (e.g., Z in FIG. 2) lose necessary flatness, thereby causing sub-wavelength random differences in optical path length through a finished wall that may produce limiting refractive and dispersive effects in acquired optical parameters; and d) the effective brittleness of the preform may increase, with increased tendency to chipping during processing of the drawn preform into prismatic flow cells or such flow cells into monolithic flow cells of the '187 patent. Each of such potential glass-forming artifacts may vary along the length of the drawn preform, so causing unit-to-unit variability in optical parameters acquired from different monolithic flow cells as well as in those acquired through individual walls of a specific flow cell (e.g., 20′ in FIG. 2). Moreover, the yield of monolithic flow cells processed in the aforesaid manner to comprise Coulter volumeter conduits may be significantly reduced by chipping at critical volumeter orifices.
The integral non-cylindrical envelope of monolithic flow cells according to the '187 patent is formed directly on at least a portion of prismatic flow cells by secondary machining processes. In addition to aforesaid glass-forming artifacts, machining artifacts may also adversely affect optical cytometric characterization parameters provided by a finished flow cell. As a first example of such artifacts, a wedge angle (i.e., a in FIG. 2) may occur in the flow-cell wall between any planar surface machined on a prismatic flow cell (e.g., 50 in FIG. 2) and the corresponding substantially planar surfaces of the prismatic channel formed in the glass-drawing operation (e.g., Z in FIG. 2). For optical flow cells including Coulter volumeter conduits as described in the '187 patent, the small conduit widths W′ (e.g., 52 micra) make controlling such wedge angle to a desired tolerance difficult during formation of the flow-cell envelope by available processes, and although the internal radius R at the joinless corners of flow channel Z is hydrodynamically advantageous, it further reduces the extent of W′ that is available as an alignment reference. As a preferable alternative, the '187 patent teaches machining a flat on the cylindrical surface of the final oversleeving tube, parallel to a planar surface of the channel, prior to drawing the preform whereby such flatted preforms have a cross-section that is substantially circular, i.e., they are substantially cylindrical. Such flats enable improved control over wall wedge angle when used as a reference during the first stages of the secondary envelope-forming processes; however, controlling the preform wedge angle between such flats and a channel surface to less than about two degrees of angle requires exceptional care, additional variability being introduced via typical flow-cell mounting techniques used during the secondary formation of envelope surfaces. While allowing acceptable coefficients of variation in lateral characterizing parameters (e.g., such as side scatter SS and fluorescence F from blood cell BC2 in FIG. 2), small variations in wall wedge angle between corresponding channel and envelope surfaces can produce disadvantageous coefficients of variation in other parameters, such as axial light loss (ALL) and low-angle forward scatter FS, that are acquired near the optical axis OA of the transducer assembly. As another example of machining artifacts, generation of a non-cylindrical surface of revolution (e.g., spherical) to provide such non-cylindrical envelope surface on a portion of a drawn preform requires such portion being centered on an axis having a desired relation to the flow-channel axis and use of, e.g., a form tool, with consequent envelope geometry depending on the accuracy and use of the forming method. Even sub-wavelength differences in optical path length between the flow channel and an envelope surface formed by secondary machining may produce refractive and dispersive effects that limit the quality of optical characterization parameters and complicate alignment of the finished flow cell during integration into a transducer assembly. Such optical artifacts originating in machining artifacts may also occur not only unit-to-unit for FIG. 2 flow cell 20′, but also for individual walls of a specific flow cell 20′.
Unacceptable optical effects of aforesaid manufacturing variability arising in both glass-forming and secondary machining can be eliminated by selection of monolithic flow cells 20′. Transducer assemblies comprising flow cells 20′ selected to provide optimum performance along FIG. 2 optical axis OA provide characterizing parameters of exceptional diagnostic quality, and certain hematology analyzers incorporate such assemblies that have survived multiple such selections. However, need for some such selection processes may not be evident until a monolithic flow cell 20′ can be functionally tested in a partial transducer assembly; rework of a flow cell 20′ or of a partial transducer assembly giving unacceptable performance adds disadvantageous costs, as well as reducing yields of both useful flow cells and transducer components.
Regardless of design, flow cells that allow acquisition of acceptable optical characterization parameters from spherical formed bodies demonstrate greater variation in such parameters acquired from formed bodies lacking at least quasi-spheroidal shape. If significantly asymmetric, such formed bodies typically transit the parameter-acquisition zone of a flow cell with their major dimension substantially aligned with the axis of sample flow, but with random orientation about said axis (i.e., the lateral profile presented at the acquisition axis is random with respect to the flow axis). Numerous formed bodies of clinical interest are asymmetric, and their random rotational orientation produces disadvantageous coefficients of variation in acquired optical characterizing parameters, with consequent increased coefficients of variation in their subpopulation data. It is known that flow channels having a rectangular cross-section preferentially orient fixed discoid red-blood cells introduced via an axisymmetric sample inlet tube (Journal of Histochemistry and Cytochemistry, 25:774-778, 1977), and this configuration has been used in flow cells for acquisition of cellular images (e.g., U.S. Pat. Nos. 5,088,816; 5,412,466; and 5,825,477). Non-axisymmetric nozzles on sample inlet tubes (e.g., as in aforesaid U.S. Pat. No. 5,825,477) and appropriately beveled tips of an axisymmetric inlet tube generate such orienting rotational forces; the latter have been used in experimental transducer assemblies to obtain S signals from oriented fixed discoid red-blood cells (Biophysical Journal, 23:1-5, 1978) or F signals from oriented stripped spermatozoa (Journal of Histochemistry and Cytochemistry, 27:353-358, 1979; Cytometry, 7:268-273, 1986). Reduced coefficients of variation in acquired optical characterization parameters, with consequent improved classification of formed bodies into subpopulations, make desirable transducer assemblies that apply such orienting rotational forces to asymmetric formed bodies in samples.
In summary, transducer assemblies comprising optical flow cells having not only aforesaid advantages of monolithic flow cells as disclosed in the '187 patent, but also enabling improved yields with less selection from production through integration into acceptably functioning transducer assemblies, would advantageously facilitate acquiring the various distinguishing parameters used by automated hematology analyzers and flow cytometers to differentiate and characterize various formed-bodies in liquid samples. Transducer assemblies comprising such flow cells having a flow-channel cross-section so formed as to apply an orienting force to formed bodies presenting asymmetric profiles when passing through the parameter-acquisition portion thereof, either separately or in combination with a sample inlet tube so formed as to apply an orienting force to such formed bodies, would advantageously decrease variability in optical cytometric characterization parameters acquired during the transductive process in such formed-body analyzers.